No longer going to be indifferent or laid-back voters, Kenyan GenZs are proving to be a fearless and arguably tribeless generation of new adults fronting a radically different worldview from their predecessors on matters of political responsibility, change, transparency, and accountability

It is time to be dispassionate and scientific! Political temperatures in Kenya should allow for rational and clear-headed decision science to help arrive at long-term solutions for shared prosperity.

Confusion Matrix is a well-known assessment tool popular with statisticians and GIS professionals. Can it help in actively involving the GenZs in decision-making?

From years of youth mentorship, this platform today shares informed insights into how decision models can be applied to engage the youth effectively in cross-generational political dialogue.

Historic and New Adventures

First published in 1970 by Oxford University Press, Moses in a Muddle could have been just another innocent title by Barbara Kimenye in her Moses series of captivating books. The author’s Ugandan origin also resonates well with the keen interest Ugandans have expressed in the recent political developments in her neighbour, Kenya. More than fifty years later, the title springs to life as the more exposed and educationally empowered GenZs modify political narratives in the East African country that has been known to be the island of peace and democracy in a troubled sea. In the timeless book, despite his best intentions as he navigates the school of life, Moses’ actions often lead to misunderstandings and mishaps, causing confusion and laughter among his peers and teachers. If the current political regime in Kenya were to personify Moses, then experimenting with ambitious taxation targets would be his school of life.

Even as GenZs have been invited to dialogue with the president, thought leaders must assert their influence for the good of the country by challenging leaders to give them room to apply scientific models that are promotive of inclusive stakeholder engagements while assuring objective and transparent decisions for shared long-term prosperity. Facing similar hurdles in sub-national revenue allocation formulae, Kenya should find the key message of this youth-centric article a resourceful guide to reaching novel and noble development goals.

The Power and Promise of a Confusion Matrix

Kenya has been thrown into a state of political confusion following the contested and finally declined Finance Bill 2024, a situation in which the GenZs (born 1996-2012) are taking control of the political agitation process using unconventional methods that are decentralised and socially connected in digitally democratised spaces, X Spaces being key. The movement has confounded Members of Parliament and diehards of traditional methods of political protests. Departing from the normal course and utilising the advantage of numbers in a country whose median age is barely 20, the new mainstream of GenZs and their youthful counterparts now steer novel, fearless, tribeless, non-linear, and issue-based approaches. Even more noteworthy is that we are witnessing interesting times in an era of big data and artificial intelligence. Convention must be challenged, exposing a regrettable decadence in the quality of expert advice availed to top decision-makers and how that explains a major part of the prevailing decision crisis and ground-shifting fiasco.

At such a time of unprecedented political protests powered by the cosmic social media revolution driven by GenZs in Kenya, questions arise on how much leaders could learn lessons in stakeholder engagement guided by tools of decision science, so as to sense political temperatures early and notice the divide in views between citizens on the ground, especially the youthful majority, and their boardroom policymakers. A ready reference tool to wade through this confusion has a fitting name, Confusion Matrix. This simple yet potent tool of data science can help evaluate the performance of any inter-party classification model, e.g., for perceptions or spatial mapping. As a table, it displays how well a model’s predictions match the actual outcomes by comparing the predicted classifications with the actual cases from ground truth.

In Context: GenZs versus Decision Makers in Kenya

In Kenya, there has lately been a noticeable gap between the perspectives of GenZs and those of politicians and decision-makers. Such a gap often leads to conflicting views on policies and proposals, particularly around cost-of-living issues such as taxation. Decision-makers generate policies based on their expert knowledge and boardroom discussions, the parliament in the case of legislators. GenZs may have different views as they are directly connected to the ground realities that hit them directly, unemployment and the high cost of living being key; they also have different priorities and shared experiences not aligned with tribe or established tradition. Using a confusion matrix, leaders can be well informed of the magnitude of the divergence to visualise and potentially resolve the stalemate before explosive protests.

Taxation Policies: An Example

Let’s take an example of how different items might be classified for taxation by decision-makers versus GenZs. Four categories for tax levels are exemplified here: zero tax, low tax, medium tax, and maximum tax. Though not a conclusive list, how decision makers and GenZs are likely to classify and assign the taxation levels is shown in the examples below.

Decision Makers’ (DM) Classification:

Zero Tax: Essential goods (e.g., basic foodstuffs, medical supplies)

Low Tax: Education materials, affordable housing

Medium Tax: Consumer electronics, luxury items

Maximum Tax: Imported diapers and similar essentials, Alcohol, tobacco, high-end luxury cars

GenZs’ (GeZ) Classification:

Zero Tax: Internet and data services, eco-friendly products, mental health services, imported diapers and similar essentials

Low Tax: Local startups, imported essentials, creative industry products

Medium Tax: Imported fashion, gadgets

Maximum Tax: Unhealthy fast food, non-essential imports

Applying the Confusion Matrix

To understand the alignment or misalignment between these two classifications, a confusion matrix might look as shown in the image.

Distributed Agreements per Tax Category

The confusion matrix demonstrates the wide departure between the views of the decision-makers and the GenZs, especially in the goods to be allocated under the low tax and maximum tax categories. At only 59.4%, the overall agreement level is low. Based on these results, the overall political leader can gauge the points of major contestation that need urgent resolution to avert a possible crisis.

Decision Makers’ Agreement per Tax Category (Vertically)

Zero Tax (DM):                7/11       =             63.6%

Low Tax (DM):                 8/18       =             44.4%

Medium Tax (DM):          15/25     =             60.0%

Maximum Tax (DM):        8/10       =             80.0%

GenZs’ Agreement per Tax Category (Horizontally)

Zero Tax (GenZ):             7/10       =             70.0%

Low Tax (GenZ):              8/12        =             66.7%

Medium Tax (GenZ):       15/24     =             62.5%

Maximum Tax (GenZ):     8/18       =             44.4%

True Positives (principal diagonal elements) are the items where both GenZ and the decision-makers agree on the classification. For example, 7 items are classified as zero tax by both. False Positives and False Negatives (Off-diagonal elements) represent disagreements. For example, 2 items that decision-makers consider to attract low tax are viewed by GenZs to attract zero tax.

Overall Agreement and Kappa

Overall Agreement

Overall Agreement = Total True Positives/Total Observations = 38/64 = 59.4%

Kappa Statistic

The Kappa statistic (0.44) = P0 – Pe /1- Pe is measures the agreement between two classifiers corrected for chance, with 1 being the perfect elimination of a random classification and -1 being a case that is much worse than a random classification. In this case, the probability of agreement by chance is Pe = 0.27 from the sum of the products of column and row totals divided by the square of 64 as the total number of categories sampled. P0 is the overall accuracy of 59.4%.

Using this analysis, decision-makers can adjust policies to better reflect the priorities and perspectives of GenZs. For example, considering zero tax on internet services and imported essentials could align more closely with GenZ’s needs, potentially fostering better engagement and compliance.

By applying a confusion matrix, stakeholders can quantitatively analyse the differences in perspectives between GenZs and decision-makers. This approach not only highlights areas of consensus but also pinpoints where adjustments are needed. In the context of Kenyan taxation policies, such a tool can bridge the gap, ensuring that policies are more inclusive and reflective of the diverse viewpoints within the society.

By Nashon Adero

Impact Borderless Digital.